Recent years have seen sharply increased and broadened use of atypical antipsychotic medications (AAPs) across the age spectrum including extensive use for non-labeled indications. Yet important basic questions remain unanswered concerning comparative safety and effectiveness of these medications, especially among vulnerable children and elderly. We will address these questions for enrollees in Medicaid, the primary mental health care payer for the most seriously mental ill. We will unite expertise and resources from three sites to address this critical public health challenge for Medicaid and Medicare beneficiaries and other stakeholders. The project consortium was organized to build on the strongest existing lines of work and relevant methodological expertise within the CERTs network, and on powerful existing merged datasets that could support the initiation of work from the start of the grant period, in order to leverage these resources into a larger and more powerful whole. Individual studies will address the following aims: [unreadable] 1. For elderly nursing home (NH) residents, utilize a unique linked dataset developed by the mental health CERT that merges medication/diagnostic information from claims with a prospective, structured clinical assessment of residents to evaluate comparative effectiveness and safety of individual antipsychotic drugs in large usual-care populations. Important clinical outcomes examined include mortality; cardiovascular, cerebrovascular, metabolic, infectious, and other medical outcomes; and functional and clinical outcomes including behavioral symptoms and cognitive function. [unreadable] 2. Assess clinical outcomes of antipsychotic therapy in large usual-care populations of children and [unreadable] adolescents. We will conduct a cohort study of important clinical outcomes and comparative safety of newly initiated AP therapy in youth (2-20 years of age) with diagnoses consistent with ADHD or behavioral or mood disorders, compared with a propensity-score-matched cohort of children with other treatment modalities. Analyses will be initially conducted in Tennessee Medicaid, with access to both electronic/paper medical records for validation of study variables/endpoints, and, using the validated definitions, extended to a national Medicaid population of youth to increase power to examine comparative safety. Among youth diagnosed with schizophrenia, we will also examine markers of effectiveness across individual drugs including drug discontinuation and psychiatric hospitalization. [unreadable] 3. For the broader population of all non-institutionalized AP users, utilize 50-state Medicaid data to determine if individual APs vary with regard to important AP risks, including prolactinomas (all ages), incident diabetes mellitus or hyperlipidemia (adults), and all-cause mortality (elderly). [unreadable] Using powerful existing merged datasets, we can initiate substantive analyses at the start of the project. Early in the project period, we will implement additional linkages and expand an existing 8-state dataset to national scope, providing enhanced power needed to assess comparative safety and effectiveness across drugs and important subpopulations. New linkages include validation of claims-based measures of adverse effects against medical charts, links between administrative and electronic medical record data, and expansion of a unique existing merged dataset on the nursing home population. The study will integrate analyses in the well established Tennessee Medicaid population with multistate Medicaid data that offer great power for comparative safety and effectiveness research. It will build on past CERTs investments, the expertise of preeminent methodological leaders, and important prior lines of research on antipsychotic use and outcomes at the three sites to implement a rigorous set of studies aimed at providing evidence that is critically needed to guide optimal clinical use of antipsychotic drugs for vulnerable populations. [unreadable] [unreadable] [unreadable]